How to use machine learning models for data prediction in FastAPI

How to use machine learning models for data prediction in FastAPI

Introduction:
With the development of machine learning, more and more application scenarios require the integration of machine learning models into actual systems. . FastAPI is a high-performance Python web framework based on an asynchronous programming framework. It provides a simple and easy-to-use API development method and is very suitable for building machine learning prediction services. This article will introduce how to use machine learning models for data prediction in FastAPI and provide relevant code examples.

Part One: Preparation
Before we start, we need to complete some preparations.

1. Install necessary libraries
   First, we need to install some necessary libraries. You can use the pip command to install libraries such as FastAPI, uvicorn and scikit-learn.

pip install fastapi
pip install uvicorn
pip install scikit-learn

2. Preparing the machine learning model
   Next, we need to prepare a trained machine learning model. In this article, we will use a simple linear regression model as an example. Models can be built and trained using the scikit-learn library.

from sklearn.linear_model import LinearRegression
import numpy as np

# 构建模型
model = LinearRegression()

# 准备训练数据
X_train = np.array(...).reshape(-1, 1)  # 输入特征
y_train = np.array(...)  # 目标变量

# 训练模型
model.fit(X_train, y_train)

Part 2: Building the FastAPI application
After the preparation work is completed, we can start building the FastAPI application.

1. Import necessary libraries
   First, we need to import some necessary libraries, including FastAPI, uvicorn and the model we just trained.

from fastapi import FastAPI
from pydantic import BaseModel

# 导入模型
from sklearn.linear_model import LinearRegression

2. Define the data model of input and output
   Next, we need to define the data model of input and output. In this article, the input data is a floating point number, and the output data is a floating point number.

class InputData(BaseModel):
    input_value: float

class OutputData(BaseModel):
    output_value: float

3. Create FastAPI application instance
   Then, we can create an instance of FastAPI.

app = FastAPI()

4. Define the route for data prediction
   Next, we can define a route to handle requests for data prediction. We will use the POST method to handle the data prediction request and InputData as the input data for the request.

@app.post('/predict')
async def predict(input_data: InputData):
    # 调用模型进行预测
    input_value = input_data.input_value
    output_value = model.predict([[input_value]])

    # 构造输出数据
    output_data = OutputData(output_value=output_value[0])

    return output_data

Part 3: Running the FastAPI application
After completing the construction of the FastAPI application, we can run the application and test the data prediction function.

1. Run the FastAPI application
   Run the following command in the command line to start the FastAPI application.

uvicorn main:app --reload

2. Initiate a data prediction request
   Use a tool, such as Postman, to send a POST request to http://localhost:8000/predict, and pass an input_value parameter in the request body.

For example, send the following request body:

{
    "input_value": 5.0
}

3. View prediction results
   You should receive a response containing the prediction results.

{
    "output_value": 10.0
}

Conclusion:
This article introduces how to use machine learning models in FastAPI for data prediction. By following the guidance in this article, you can easily integrate your own machine learning model into your FastAPI application and provide prediction services.

Sample code:

from fastapi import FastAPI
from pydantic import BaseModel
from sklearn.linear_model import LinearRegression
import numpy as np

# 创建模型和训练数据
model = LinearRegression()
X_train = np.array([1, 2, 3, 4, 5]).reshape(-1, 1)
y_train = np.array([2, 4, 6, 8, 10])
model.fit(X_train, y_train)

# 定义输入输出数据模型
class InputData(BaseModel):
    input_value: float

class OutputData(BaseModel):
    output_value: float

# 创建FastAPI应用实例
app = FastAPI()

# 定义数据预测的路由
@app.post('/predict')
async def predict(input_data: InputData):
    input_value = input_data.input_value
    output_value = model.predict([[input_value]])
    output_data = OutputData(output_value=output_value[0])
    return output_data

I hope that through the introduction and sample code of this article, you can successfully use machine learning models for data prediction in FastAPI. I wish you success!

The above is the detailed content of How to use machine learning models for data prediction in FastAPI. For more information, please follow other related articles on the PHP Chinese website!